Documentation for population



Nutrient Budgeter User Guide

Sylvia C. Schaefer and Merryl Alber

Department of Marine Sciences, University of Georgia, Athens, GA

Table of Contents

Organization & Overview 4

General information 5

Getting Started 5

Population module 8

Livestock Module 9

Crop Module 14

Net Food and Feed Import 22

Fertilizer Module 23

Forest Module 25

Atmospheric Deposition Module 26

Total Inputs 27

Output 27

Input Variables: Note to Mac Users 29

Additional Information 30

Acknowledgments

We thank Adrian Burd and Joan Sheldon for assistance with Matlab programming.

Introduction

This program calculates the input components of a SCOPE-style nitrogen budget (e.g. Boyer et al. 2002) and a similar phosphorus budget (e.g. Schaefer and Alber 2007a) from county-level input data. We suggest you consult one of these papers to further familiarize yourself with the methodology behind these budgets.

Most of the modules are optimized for use with the particular sources of data used in those budgets, but can be used with other data as well. Creative users will easily find additional applications for this code—while the dialog boxes may ask you for annual N consumption rates, there is nothing preventing you from entering the annual kumquat consumption rate instead if watershed kumquat consumption is more your cup of tea! Some of the modules calculate multiple components of the budgets; if there are some you are not interested in, simply enter 0 when prompted for rates related to that component.

The program was developed primarily to deal with data from the United States Census of Agriculture, which frequently withholds values to protect the privacy of farmers. Details on the manner in which these values are estimated are described in the sections on livestock and crop modules.

Running this code requires access to Microsoft Excel and to a copy of Matlab with the Statistics Toolbox. The program runs on both Windows and Apple machines; however, a Windows system is needed for Matlab to correctly read information from Excel. If you are running this program on a Mac, please read the special note to Mac users at the end of this manual.

Organization & Overview

The nitrogen inputs and exports calculated in this program are as follows:

Inputs Exports

Net Food and Feed Import Non-Food Crop Export

Fertilizer Use Manure Volatilization

Agricultural N Fixation Fertilizer Volatilization

Forest N Fixation

Atmospheric Deposition

The phosphorus inputs and exports are as follows:

Inputs Exports

Net Food and Feed Import Non-Food Crop Export

Fertilizer Use

The program is organized into six modules, all of which can be run independently of one another, and which calculate the following components:

Population Module: Watershed population

Human consumption (Net food and feed import)

Livestock Module: Livestock consumption (Net food and feed import)

Livestock excretion (Net food and feed import)

Livestock production (Net food and feed import)

Manure N volatilization

Crop Module: Crop production (Net food and feed import)

Agricultural N fixation

Non-food crop export

Forest Module: Forest N fixation

Fertilizer Module: Fertilizer use

Fertilizer N volatilization

Atmospheric Deposition Module: Atmospheric N deposition

Net food and feed import can only be calculated if the population, livestock, and crop modules are all run. See the net food and feed import section of this manual for more information.

General information

The U.S. government has assigned each state and county within the U.S. a FIPS (Federal Information Processing Standard) code, and these are used for most forms of federal county-level data. These modules were written with these codes in mind; however any type of numeric code may be used. The only restriction is that state (or equivalent) codes should be divisible by 1000 so that the program will recognize these as state, rather than county, totals.

Please note that throughout, you are responsible for ensuring that your values are in the correct units. For example, if you input acreages to the program, please double-check that your related variables, such as crop yields, nitrogen fixation rates, etc. are in per-acre units.

Data are read in from Microsoft Excel spreadsheets set up as shown in the module descriptions below. These spreadsheets should be placed in the same directory as the program files. We have provided these spreadsheets containing sample data; replace these sample data with your own when you are ready to calculate inputs for your own watersheds.

Getting Started

Two input files are used by all modules:

input_watershed_areas.xls

Contains the watershed names and areas. Set up the file with one row containing watershed names and the next containing the corresponding areas:

Watershed 1 name Watershed 2 name Watershed 3 name …

Watershed 1 area Watershed 2 area watershed 3 area …

For example,

[pic]

If you would prefer total inputs, rather than per-unit-area inputs, set these areas to 1. Clearly, you are not restricted to watersheds, but could use any geographical unit of interest.

input_watershed_fractions.xls

Contains the fraction of each county (or equivalent unit) located inside the watershed. The file should be set up as follows:

FIPS Codes Watershed 1 Watershed 2 Watershed 3 …

Code 1 Fraction Fraction Fraction

Code 2 Fraction Fraction Fraction

Code 3 Fraction Fraction Fraction

. . . .

. . . .

. . . .

For example,

[pic]

If a cell has been left blank in this file, or if the program cannot find corresponding fractions for a particular county code when performing calculations on data, it will assume 0. In some cases, you may wish to use a different set of fractions for certain modules—for example, if you prefer to use census tract-level information for your human population and consumption estimates, or percentages of agricultural rather than total land area for the agricultural components of your budgets. In such a case, you will need to run the program separately with each set of watershed fractions and add the numbers manually at the end.

Please make sure that the watersheds are in the same order in both files.

To start the nitrogen budgeter, open Matlab and type nutrientbudgets at the prompt. You will see a welcome screen which prompts you to select the modules and the nutrients you wish to run. Click “Get Started!” when you are ready to begin. The modules will be run consecutively.

[pic]

Population module

The population module calculates human population and N and/or P consumption in a watershed. It requires one additional input file:

input_population.xls

Contains the raw population per county (or other unit) of interest. It should be set up as follows:

County Code Population estimate

Code 1 County 1 estimate

Code 2 County 2 estimate

Code 3 County 3 estimate

. .

. .

. .

The “FIPS Code” and “Population estimate” headers are optional in this file.

For example,

[pic]

The program asks for only one additional piece of information, namely the per-capita consumption rate, and then returns the watershed population and per-unit-area human consumption. The defaults are 5 kg N person-1 year-1 (Garrow et al. 2000) and 0.5475 kg P person-1 year-1 (Smil 2000):

[pic] [pic]

Livestock Module

The livestock module calculates livestock N and/or P consumption, production, and excretion, and manure volatilization. It requires one additional input file:

input_livestock.xls

Contains the livestock populations in each county. The county values in each state should be preceded by the state total. Setup is as follows:

FIPS Code Animal 1 Animal 2 Animal 3 …

State 1 Code State total State total State total

County 1.1 County value County value County value

County 1.2 County value County value County value

County 1.3 County value County value County value

. . . .

. . . .

. . . .

State 2 Code State total State total State total

County 2.1 County value County value County value

County 2.2 County value County value County value

County 2.3 County value County value County value

. . . .

. . . .

. . . .

For example,

[pic]

Note that in this example, 13000 is the code for the state total (in this case, Georgia), and codes starting with 13001 are county codes. If you are not using U.S. FIPS codes and your data will require separation by state (or equivalent) in order for the missing values to be filled in (as described below), please still assign your state totals a code divisible by 1000, as this is how the program recognizes the beginning of a new state. Please note that if the last column (last form of livestock) of your input spreadsheet contains no data, this may cause an error. In such a case, please go ahead and delete the last column.

If this module is to be used with data from the U.S. Agricultural Census, many data values may be denoted with a “(D)” code. This indicates that the data were withheld to protect the privacy of farmers in that county. Because Matlab is unable to handle both text and numerical input values in a dataset, these values must be changed to -999 in the spreadsheet in order for the module to handle these withheld values properly. We suggest autofiltering the raw data in Excel to change all (D) values to -999, then using the pivot table features to easily achieve this layout. Add the county codes in the Row Fields, livestock types in the Column Fields, and the livestock numbers to the data items.

Withheld values are calculated by the module as:

[pic]

On a PC system only, after calculating these values, the module writes an output file output_livestock.xls. This file contains the raw county data, with “Withheld” codes replaced by estimated values in those counties where data were withheld. The file is intended for export to a database such as Microsoft Access, and is set up as follows:

County Code 1 Value Crop Name

County Code 2 Value Crop Name

County Code 3 Value Crop Name

. . .

. . .

. . .

The program will then calculate the number of each type of livestock in a watershed based on the watershed_fractions.xls file.

The module will ask if you need to perform an estimation for young beef and milk cattle:

[pic]

If you are using U.S. Agricultural Census data, you may wish to use this portion of the module. The U.S. Agricultural Census reports numbers of Milk Cows, Beef Cows, and Total Cattle and Calves in each county. If you select “Yes,” the program will calculate the number of young beef and milk cattle in a watershed as, respectively

[pic]

and

[pic]

This calculation cannot be done on a county level when setting up your initial spreadsheet because some of the county values may be withheld. You will need to input the column numbers for Total Cattle, Milk Cows, and Beef Cows:

[pic]

The FIPS codes do not count as column numbers, so in the example spreadsheet shown above, the column numbers would be 1, 2, and 3, respectively. These are the defaults in the dialog box. It is recommended that you move these items to the first few columns of your spreadsheet in order to simplify entering these numbers.

You will then be asked for the consumption, excretion, and manure volatilization rates for each type of livestock.

[pic]

Clicking the “Suggestions” button will bring up tables of suggested values:

[pic]

[pic]

Since most livestock emissions are redeposited locally, you will be asked what fraction of volatilization you would like to consider as a long-range transport. This fraction is considered an export from the watershed.

[pic]

0.25 is the default value used in the SCOPE Project.

The program will then calculate total consumption, total excretion, total production (the difference between consumption and excretion, minus 10% to account for spoilage and loss), and manure volatilization export.

Crop Module

The crop module calculates crop production, crop N fixation, and non-food crop export. Depending on your needs, it requires between one and three input files.

input_crop.xls

This is the only required input file. It contains areas and harvested quantities for each type of crop on alternating lines. This format is easily set up using Pivot Tables in Microsoft Excel. Leave the Code column blank for lines containing harvest. The county values for each state should be preceded by the state totals:

County Code Crop1 Crop2 Crop3 …

State 1 Code area area area

harvest harvest harvest

County 1.1 area area area

harvest harvest harvest

County 1.2 area area area

harvest harvest harvest

County 1.3 area area area

harvest harvest harvest

. . . .

. . . .

. . . .

State 2 Code area area area

harvest harvest harvest

County 2.1 area area area

harvest harvest harvest

County 2.2 area area area

harvest harvest harvest

County 2.3 area area area

harvest harvest harvest

. . . .

. . . .

. . . .

For example,

[pic]

Please make sure that every other row of the county code column is blank, rather than containing the county code again. Again, if you will not be using U.S. FIPS codes and your data will require separation by state (or equivalent) in order for the missing values to be filled in (as described below), please still assign your state totals a code divisible by 1000.

Values withheld should be changed to -999. Again, it is recommended that you autofilter the data in Microsoft Excel to change “(D)” values to -999 and then apply the pivot table functions. Change any other non-numeric values to 0 or leave blank. Using Microsoft Excel’s Pivot Tables feature, add the county codes in the Row Fields, crop names in the Column Fields, and area and then quantity to the data items. Copy the data to the input spreadsheet and remove any unnecessary rows/columns. Please note that if the last column (last crop) of your input spreadsheet contains no data, this may cause an error. In such a case, please delete the last column.

The program fills in the missing values by running a linear regression between area and harvested quantities for each state. These regressions tend to be highly significant, often with R2 values approaching 0.99. If the regression is significant at the p ≤ 0.05 level, items missing either area or harvested quantity (but not both) are filled in using the equation of the regression line. If both values in a county are missing, the regression is not significant, or there were insufficient data for that crop to even attempt a regression, the remaining missing values are filled in as for livestock.

The module will ask you if you have any crop data for which only areas, but not harvested quantities, were reported:

[pic]

In the U.S. Agricultural Census, this is common for crops such as vegetables and orchard crops, and you will have an opportunity to provide estimated yields for these crops so that they can be included in the N and/or P production calculation. You may set up a second spreadsheet containing these values:

input_crop_no_harvest.xls

This spreadsheet will contain only the areas for each crop, set up similarly to the input spreadsheet for the livestock module:

FIPS Codes Crop1 Crop2 Crop3 …

State 1 Code area area area

County 1.1 area area area

County 1.2 area area area

County 1.3 area area area

. . . .

. . . .

. . . .

State 2 Code area area area

County 2.1 area area area

County 2.2 area area area

County 2.3 area area area

. . . .

. . . .

. . . .

For example,

[pic]

The program will calculate values missing in this spreadsheet (as usual, denoted by -999) as for livestock; i.e.

[pic]

In order to determine which crops are sufficiently important in a watershed to merit inclusion in calculations of crop production, you will be asked what fraction of harvested cropland you wish to use as a cutoff for including a particular crop in your calculations. 0.01 (1%) is a recommended value:

[pic]

You will then be asked if you have data on the area of total harvested cropland in each county. This is used in the calculation of which crops meet the cutoff criteria you entered above.

[pic]

If you do not have areas of harvested cropland, click “No” and the program will sum areas for all the crops that were entered. If you do, this should be set up in a separate spreadsheet:

input_harvested_cropland.xls

This spreadsheet should contain state and county codes, and state and county totals of the areas of harvested cropland:

Code Harvested Cropland

State 1 Code area

County 1.1 area

County 1.2 area

County 1.3 area

. .

. .

. .

State 2 Code area

County 2.1 area

County 2.2 area

County 2.3 area

. .

. .

. .

For example,

[pic]

The headers are optional.

Although there are unlikely to be any, the program will check for withheld values (which again should have been set to -999) and estimate as for livestock.

If you entered crops without associated data on quantity harvested, the program will then ask you for estimated per-unit-area yields for these crops. You will only be asked about crops consisting of more than your specified fraction of harvested cropland. Remember to enter the yield in the correct units (for example, enter pounds per acre if your areas are in acres):

[pic]

The program will estimate harvest in each watershed based on the entered yield.

Hint: A good reference for estimated yields is the California Department of Agriculture’s Agricultural Resource Directory.

Hint: You can enter “0” as a yield if you do not want a particular crop included in the calculation, for example because it is a subset of another crop. For example, you may be asked for yields for both “Pecans, All,” and “Pecans, Improved” if both these were included in your initial spreadsheet. You may wish to enter “0” as the yield rate for “Pecans, Improved” to avoid double-counting a crop. Or delete it from your spreadsheet before running the module—that would be better!

You will be asked if you wish to calculate non-alfalfa hay:

[pic]

This is a calculation that was used for nitrogen budgets using the SCOPE methodology (Boyer et al. 2002), and consists simply of subtracting alfalfa hay areas and yields from total hay areas and yields. However, since alfalfa hay data are frequently withheld, this cannot be done as part of spreadsheet set-up. If you wish to do this calculation, you will be asked which spreadsheet hay and alfalfa hay were in (with or without associated harvest data), and the column numbers for the two crops:

[pic]

As with the milk/beef cattle calculation in the livestock module, do not count the first column of FIPS codes when entering column numbers. Again, it is recommended that you make these crops the first two columns in your spreadsheet if you intend on doing this calculation in order to simplify entering the column numbers.

On a PC system only, the program writes out the raw areas and harvest data for each county to a file, output_crop.xls. Withheld values are replaced by the estimated values. This output is designed for subsequent import to a database such as Microsoft Excel and is formatted as follows:

County 1 Code Area Harvested quantity Crop name

County 2 Code Area Harvested quantity Crop name

County 3 Code Area Harvested quantity Crop name

. . . .

. . . .

. . . .

You will then be asked several questions about each crop that meets your specified criteria for amount of harvested cropland:

[pic]

Conversion rate: This is the amount of N (or other element of interest) per unit harvested. Remember to check your units.

Hint: As with yield, you may enter “0” as the conversion rate if you do not want a particular crop to be included, for example to avoid double-counting. Alternately, remove such crops from your spreadsheet before running these modules!

N fixation rate: The amount of N fixed per unit area. Enter 0 if the crop is not a nitrogen fixer. Or, you may instead enter some other per-area rate of interest, although your results will still be labeled as “N fixation.” Remember to check your units.

Forage/silage crop: For most crops, 10% of crop production is subtracted to account for spoilage or other loss. However, this is not done for forage and silage crops.

Non-food crop: You may wish to exclude some crops from inclusion in the net food and feed import calculations (for example, cotton and tobacco, which are not food crops). Crops for which you indicate that they are non-food are not included in total crop production, but tallied separately as a non-food crop export.

Again, clicking the “Suggestions” button will bring up a window with some suggested values:

[pic]

Net Food and Feed Import

If you run all three of the population, livestock, and crop modules, the program will automatically calculate net food and feed import. This is defined as total consumption minus total production, or (human consumption + animal consumption) – (crop production + animal production). It can be thought of as the amount of food and feed required to sustain the populations in a watershed above and beyond that which is already produced in that watershed.

If you do not run all three modules, those components of net food and feed import which you do calculate will still be reflected in the final budget. Human and animal consumption will are treated as an import; crop and animal production are treated as an export.

Fertilizer Module

The fertilizer module calculates fertilizer use and volatilization from county sales data. Potential sources of this information for the U.S. include Battaglin and Goolsby 1994 and Ruddy et al. 2002. The data should consist of the weight of N in fertilizer in each county, rather than the total weight of the fertilizer. The input files input_Nfertilizer.xls and input_Pfertilizer.xls (for nitrogen and phosphorus fertilizer forms, respectively) contain county codes and amount of fertilizer used in each county:

County Code Type1 Type2 Type3 …

County 1 quantity quantity quantity

County 2 quantity quantity quantity

County 3 quantity quantity quantity

. . . .

. . . .

. . . .

For example,

[pic]

For nitrogen fertilizers only, the module will then ask for a volatilization rate for each type of fertilizer. A table of suggested values pops up automatically.

[pic]

Finally, since most volatilization is re-deposited locally, you will be asked what proportion of volatilization you wish to consider as an export:

[pic]

0.25 is the default that was used in the SCOPE Project.

Fertilizer use (both N and P) in the watersheds is calculated automatically, without any additional input, by (as usual) multiplying the nutrient in fertilizer in each county by the fraction of the county in the watershed.

Forest Module (N only)

The forest module calculates forest fixation and requires input_forest.xls, containing county codes and area of forest type. This should be set up as follows:

County Code Type1 Type2 Type3 …

County 1 area area area

County 2 area area area

County 3 area area area

. . . .

. . . .

. . . .

For example:

[pic]

For each forest type, you are then asked for the N fixation rate and the proportion of area you wish to apply it to:

[pic]

For example, if you wish to assume, as in SCOPE budgets, that black locust makes up 10% of oak/hickory forest and fixes N at a rate of 5,000 kg N km-2 yr-1, enter 5000 and 0.1 in this dialog box.

Atmospheric Deposition Module (N only)

If you have county-level atmospheric input data, such as that in Ruddy et al. 2006 (the source of the data in the sample file), you can run the atmospheric deposition module. This module calculates watershed atmospheric deposition and requires one input file, input_atmospheric.xls, which should contain county codes and atmospheric inputs:

County Code Atmospheric Input

County1 Input

County2 Input

County3 Input

. .

. .

. .

Headers are optional in this file.

For example,

[pic]

No further information is required from the user to calculate watershed totals.

Total Inputs

The program calculates the total inputs from the modules which you have run. The total N input is calculated as

(net food and feed import + crop fixation + forest fixation + fertilizer use + atmospheric deposition) – (fertilizer volatilization + manure volatilization + non-food crop export)

where

net food and feed import = (human consumption + livestock consumption) – (crop production + livestock production)

For P inputs, the calculation is:

(net food and feed import + fertilizer use) – (non-food crop export)

If you have not run all the modules, those components which were not calculated are considered to be 0 in this calculation.

Output

After the program has run all the modules, you will be taken to a screen where you can view each component as a bar graph. The bars are labeled with the first three letters of each watershed’s name. Select your desired budget component from the drop-down menu and click “Update” to see the results. The drop-down menu includes all components of a complete nitrogen budget; if you have not run all of these modules then the components you did not calculate will display as 0 for all watersheds.

[pic]

The components that were calculated are also written to N_output_budget.xls and/or P_output_budget.xls, and, in comma-delimited format, to N_output.txt and/or P_output.txt. On a Mac, only the text files are written. These files are overwritten with each run of the program, so please save the data to another file before running the program again if you wish to retain your results from a previous run. Please note that in the Excel file, Matlab will overwrite only enough cells to account for that run’s output data, so if you ran all modules and then rerun only one of them, you will still see output from the first run in your Excel file unless you deleted that data before rerunning the program.

Input Variables: Note to Mac Users

Unfortunately, the Mac version of Matlab does not do a good job of reading data from Excel files, and you will have to set up the inital variables differently and in some cases by hand.

The numeric data in the spreadsheets should be exported as comma-delimited files with the same name as the .xls spreadsheets described above, but with a .csv extension. For example, instead of creating fertilizer_input.xls, you would create fertilizer_input.csv and place it in the same directory as the program files. However, please remove the text headers from these files. You will need to set the variables containing headers up manually:

cropnames

Cell array containing crop names (in the same order as the data in the crop input file). Required for the crop module.

Example: cropnames = {'Corn' 'Wheat' 'Oats' 'Soybeans' 'Peanuts'};

fertilizer_forms

Cell array containing the names of fertilizer forms (in the same order as the data in the fertilizer input file). Required for the fertilizer module.

Example: fertilizer_forms={'Anhydrous ammonia' 'Aqueous ammonia' 'Ammonium nitrate' 'Ammonium sulfate' 'Nitrogen solutions' 'Urea' 'Other'};

forest_types

Cell array containing the names of forest types (in the same order as the data in the forest input file). Required for the forest module.

Example: forest_types={'Total' 'White oak / red oak / hickory' 'Overcup oak / water hickory'};

livestock_names

Cell array containing the names of livestock types (in the same order as the livestock input file). Required for the livestock module.

Example: livestock_names={'Total cattle' 'Beef cattle' 'Milk cattle' 'Hogs' 'Broiler chickens' 'Layer chickens'};

noharvest_cropnames

Cell array containing the names of crops without associated harvest data (in the same order as the input file for crops without associated harvest data). Optional for the crop module; create only if you are entering crop area data without associated harvest data (see the module description for more information).

Example: noharvest_cropnames={'Apples' 'Asparagus' 'Brussels Sprouts' 'Rutabagas' 'Zucchini'};

watershed_names

Cell array containing watershed names (in the same order as the watershed area input file). Required for all modules.

Example: watershed_names={'Altamaha' 'Black' 'Cape Fear' 'Edisto' 'Neuse' 'Ogeechee' 'Pamlico' 'Pee Dee' 'Roanoke' 'Santee' 'Satilla' 'Savannah'};

Save these variables into a workspace in the same directory as the program files and name it mac_header_arrays.mat before running the program.

Additional Information

The livestock and crop modules use relatively complex data structures to store and process the agricultural data. For advanced Matlab users who are interested, the data structures are organized as shown below.

livestock data structure, contains data organized by state, then by crop

[pic]

livestocktotals, contains watershed totals organized by crop

[pic]

crops, contains data organized by state, then by crop

[pic]

croptotals, contains watershed totals organized by crop.

[pic]

References

Battaglin WA, Goolsby DA (1994) Spatial Data in Geographic Information System Format on Agricultural Chemical Use, Land Use, and Cropping Practices in the Unived States. USGS Water Resources Investigations Report 94-4176 Report available online at .

Battye R, Battye W, Overcash C, Fudge S (1994) Development and selection of ammonia emission factors. Final Report prepared by EC/R Incorporated for EPA Atmospheric Research and Exposure Assessment Lab, EPA Contract Number 68-D3-0034

Boyer EW, Goodale CL, Jaworski NA, Howarth RW (2002) Anthropogenic nitrogen sources and relationships to riverine nitrogen export in the northeastern U.S.A. Biogeochemistry 57:137-169

CDFA; California Department of Food and Agriculture (2009) California Agricultural Resource Directory 2008-2009. URL:

Garrow JS, James WPT, Ralph A (Eds) (2000) Human nutrition and dietetics 10th ed. Churchill Livingstone, Edinburgh, 900 pp

Ruddy BC, Lorenz CL, Mueller DK (2006) County-level estimates of nutrient inputs to the land surface of the conterminous United States, 1982-2001. USGS Scientific Investigations Report 2006-5012, Reston, VA

Schaefer SC, Alber M (2007a). Temporal and spatial trends in nitrogen and phosphorus inputs to the watershed of the Altamaha River, Georgia, USA. Biogeochemistry. 86(3):231-249.

Schaefer SC, Alber M (2007b) Temperature controls a latitudinal gradient in the proportion of watershed nitrogen exported to coastal ecosystems. Biogeochemistry 85:333-346

Smil V (1999) Nitrogen in crop production: an account of global flows. Global Biogeochem Cycles 13: 647-662

Smil V (2000) Phosphorus in the environment: Natural flows and human interferences. Ann Rev Energ Env 25: 53-88

USDA/NASS; U.S. Department of Agriculture, National Agricultural Statistics Service (2007) 2007 Census of Agriculture. URL:

USDA/NRCS; United States Department of Agriculture—Natural Resources Conservation Service (2009) PLANTS Database Crop Nutrient Tool. Data available online at

Van Horn HH (1998) Factors affecting manure quantity, quality, and use. Proceedings of the Mid-South Ruminant Nutrition Conference, Dallas-Ft. Worth, May 7-8, 1998. Texas Animal Nutrition Council, pp 9-20

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download